seisnn.model.evaluator.GeneratorEvaluator

class seisnn.model.evaluator.GeneratorEvaluator(database=None, model_name=None)[source]

Bases: seisnn.model.evaluator.BaseEvaluator

Trainer class.

__init__(database=None, model_name=None)[source]

Initialize the evaluator.

Parameters

  • database – SQL database.

  • model_name – Saved model.

Methods

__init__([database, model_name])

Initialize the evaluator.

get_dataset_length()

get_eval_dir(dir_name)

get_model_dir(model_instance)

predict(tfr_list[, batch_size])

Main eval loop.

score(tfr_list[, batch_size, delta, height, …])

Attributes

database

predict(tfr_list, batch_size=500)[source]

Main eval loop.

Parameters

  • tfr_list – List of .tfrecord.

  • name (str) – Output name.